La Mano Dashboard
The vision system sees. The dashboard translates. The owner decides.
La Mano Vision produces a detailed event log for every production session — every appearance, packaging event, and exit, timestamped to the frame. That data is exactly what you need for debugging the tracker. It is not what you need to tell the owner how many bags were made today.
The dashboard closes the loop. La Mano Vision is the brain. The dashboard is the interface the business actually uses.
Raw CSV output requires a developer to interpret it. The store owner needed the answer to a single question — how many bags did we make today — without opening a terminal or reading a log file.
Beyond the daily count, the owner wanted to see production by hour (to understand when the team is most productive), and historical trends by day and week. These are the questions any production manager asks. The system just needed to answer them from data it was already capturing.
The dashboard reads from the events.csv output that La Mano Vision writes after each session. It parses the event log, filters for completed bag events, and renders:
Total completed bags for the current session, updated each time a new events.csv is generated.
Bags completed per hour — useful for spotting productivity peaks and slowdowns across a shift.
Daily and weekly totals over time, so the owner can compare this week to last week without any manual tracking.
The dashboard is a Next.js 15 app with static generation at build time and a lightweight API route that reads the latest events.csv from the shared output directory. No database — the CSV is the source of truth because La Mano Vision already writes it reliably.
Choosing CSV over a database was a deliberate trade-off. A database would be more queryable and scalable. It would also be another thing to maintain, backup, and explain to a non-developer. The CSV approach means the entire pipeline can be understood by reading two files: the tracking script and the dashboard parser. That matters when the person responsible for keeping it running isn’t a developer.
Deployed on Vercel. The owner accesses it from any device — phone, tablet, or the counter laptop — with no setup required.
La Mano Vision and La Mano Dashboard are one system in two parts. The vision pipeline handles the hard problem — detecting, tracking, and counting bags from raw camera footage. The dashboard handles the communication problem — making that count legible to the person who needs to act on it.
Neither is the interesting part on its own. The interesting part is that a family business now has real production data, updated daily, without any manual counting and without any developer needing to be present for it to work.